Hello, everyone. Welcome to our earnings call today. Before the, before we report the financial numbers of this quarter, we would like to make an announcement for the reorganization of our structure. As you can see from the screen, the, the slide. Well, as the EV business has grown in significance over the past few years, we've decided to separate it from the power electronics segment. That is to say, we are restructuring our existing business categories, moving from three segments, power, which are power electronics, automation, and infrastructure, to four major segments, which are power electronics, mobility, infrastructure, and automation. This new organizational structure will take effect at the beginning of next year. As usual, we will have our IRO, Rodney, to report, to present the Q3 financial results.
So meanwhile, I would like to thank you for voting us in the IR, the Institutional IR Awards. So now we are going to review the financial number of Q3. So Q3 revenue was TWD 107.8 billion, representing a 1% year-on-year growth and a 7% sequential increase. So the sales revenue was seasonal, but slightly below our initial expectations. So GP in Q3 grew 9% quarter-on-quarter, but moderated by 1% year-on-year, which was actually in line with the expectations, because the comparison base from last year was actually relatively high.
So thanks to the improved scale, GP margin in Q3 increased to 29.6% from 29.2% in Q2, but dropped from the high base of 30.3% a year ago, due to the disadvantage product mix. So Q3 expense was up 7% year-on-year and 5% quarter-on-quarter. As a result of wage increases for engineers, the R&D expense grew faster than the SG&A. So the OpEx has grew by 7% compared to the previous year. So as a result, Q3 operating profit was up 16% quarter-on-quarter, but down 12% year-on-year from the high base of last year.
The OP margin improved to 11.0% versus 10.2% in Q2, but dropped from the high base of 12.7% a year ago. The performance by segment varies from one segment to another. For example, as a combined result of lackluster macro demand and a high comparison base from the previous year, we saw revenue contractions in Q3 for several businesses, including IA, BA, telecom power, networking, and EV charger. While the EV component business, server and server or data center-related power supply business, energy storage system business continued to see decent growth in Q3. For the net operating profit, Q3 net operating profit was TWD 1.9 billion , slightly increased from TWD 1.6 billion in Q2.
The foreign exchange income provided the most the most incremental gains, thanks to the strength in the U.S. dollar in Q3. In Q3, we had TWD 13.8 billion profit before tax, up 15% quarter-on-quarter, but down 9% year-on-year. EBITDA was TWD 19.7 billion, up 13% quarter-on-quarter, but down 2% year-on-year. Q3 tax expense was about TWD 2.7 billion, representing a 20% effective tax rate. As a result of Delta Thailand's earnings growth, the non-controlling interest has increased in Q3 from Q2 and a year ago. So Q3 net profit after tax was about TWD 9.3 billion, up 15% quarter-on-quarter, but down 15% year-on-year. So the EPS in Q3 was 3.60.
Now we have a look at the accumulated numbers for the first three quarters. The revenue was TWD 301.2 billion in the first three quarters, up 8% from a year ago. Gross profit was up 7% year-on-year, while GP margin was 28.8% versus 29.1% a year ago. The operating expense in the first three quarters was up by 11% year-on-year, with SG&A up 10% and R&D up 13%. OpEx ratio increased to 18.7% from 18.1% a year ago, with the SG&A expense ratio expanding to 10.0% from 9.8% a year ago. And the R&D expense ratio increased to 8.7% from 8.3%.
So the OP was flattish compared to the previous year, and OP margin in the first three quarters was 10.2% versus 11% a year ago. So, in terms of the performance by segment, we found decent revenue growth in power electronics, a modest and more from a load increase in automation, and a flattish growth in the infrastructure segment. So for the earnings, except power electronics, so strong profit improvement, the profits of automation and infrastructure shrank from a year ago due to the high base and slow economy. Therefore, in our operating profits, we had about TWD 5.1 billion in our operating profit in the first three quarters. In total, we had TWD 35.7 billion pre-tax income, up 3% from a year ago.
Our EBITDA was TWD 52.4 billion, up 7% from a year ago. So the effective tax rate was, for the first three quarters, 20%, which has been quite stable. And therefore, the net profit after tax in the first three quarters was TWD 24.4 billion, versus TWD 24.8 billion. So the EPS in the first three quarters of the year was 9.40 versus 9.53 a year ago.
So now we are open to your questions.
So the first question here is regarding the auto market. So in the recent earnings call of Tesla, Panasonic, and ON Semiconductor, concerns about softer demand for electric vehicles were mentioned. So have you seen any signs of slower demand for electric vehicles?
So I think, due to the impact of the high-interest rate environment, the current macro environment is indeed somewhat unfavorable to the whole auto market. So, and then, as for the strike by the U.S. auto workers, has been somewhat, I mean, the strike. I think the U.S. auto workers have already reached some preliminary agreements with their companies. So in my point of view, I think the electric vehicle market will continue to grow quite decently, I mean, in the coming years. But speaking of the price of the electric vehicles, I do believe there is actually room for to... I mean, room for further reduction.
And if you look at the price, the prices of the batteries, they have also been coming down, I mean, from the high base in the previous, and they have also been coming down recently, and that is going to be helpful on the price reduction of the whole vehicles. So if you are, I mean, curious about whether the strike in the U.S. had any kind of impact or has any kind of impact on our business, I think, to some extent, we did see some impact from this strike. But I think, besides the strike in the U.S., there are also many other uncertainties in the market, including the wars between and the geopolitical tensions among many countries.
Then also, the petrol price is another critical factor to the inflation. Before we recognize the [audio distortion], how, which are... So, considering the likelihood of increasing continually, continuously increasing petrol price, I think that is going to be beneficial to the auto market. Because after all, if you set, put aside the concerns on the environment, still, as a consumer, what really matters is the price of the whole vehicle. So I think that is going to be, I mean, one of... it is going to be helpful for the auto market.
So the next question would be, the outlook for Q4 and for the next year, what are the main growth drivers, and what is the trend in revenue and gross margin?
So in terms of sales momentum, I think the sales momentum in Q4 is going to be similar with Q3. And for next year, we do expect some recovery for the businesses across many segments. Are there any signs of demand recovery in consumer electronics? I think it's difficult to say that we have already seen a recovery. But it hasn't really returned, because it hasn't really returned to positive growth after all. However, compared to before, the pace of decline has indeed significantly slowed down. So next year, I think it's going to be better, I mean, with the low base of this year.
So could you please, I mean, comment on the revenue contributions, order visibility, capacity status, future growth prospects, and growth margin range for the AI server power supplies?
I think if you look at the hyperscalers, the major hyperscalers, they are all in fierce competition in building up their capability in the high-performance computing as well as the AI technology. So, but if you ask me, like, about the sales contribution, to our overall revenues, I think it's hard to estimate because as I explained before, even though they are—I mean, the hyperscalers, they are in fierce competition in this, AI technology development, but still, they, there are, they still lack the business models of AI applications. So whether or not, I mean, they can or we as a supplier, can translate, their ambitions in the AI development to the real shipments of the products or revenues, is, is, is hard to, hard to make a forecast.
So for the AI technology, it's pretty costly, not just on the CapEx level, but also on the OpEx level, which means that the running costs of the AI models is highly costly to those hyperscalers. And then also the AI. I mean, in order to run the AI models, it can be hugely, it can be, I mean, super power hungry. I mean, those equipment are super power hungry in order to run the AI training models.
So at the end of the day, what I was trying to say is, even though that we are really optimistic about the development of the AI trend, but still, I mean, coming back to whether or not we can, I mean, see mature business models, I mean, in the market in the near term, which can... Because if they don't, I mean, for the customers, if they don't find the applications or the business models, those investments couldn't be sustainable. So in short, so, I think the AI, I mean, the, the AI demand is definitely going to be one of the main growth drivers for the company in the longer run.
So my next question is, can you please give us more colors on the revenue contribution and outlook for the cooling solutions of AI servers?
So we actually provide various cooling solutions, including air cooling and liquid cooling. So I think, if you look at the heat, I mean, the density for the AI servers, the cooling technology is going to be, I mean, play a vital roles in the AI development. So I think that we will continue to work on this.
So my next question is, what are the conditions for the recovery of IA demand in China? When can we expect a more significant recovery?
So I think the IA market in China this year has been somewhat lackluster. But still, considering the labor shortage and many other factors in the China market, that is, I mean, at some point, we still think that this China IA market is going to recover—is going to recover. But just adding, I mean, in the near term, we haven't really seen any significant or any, I mean, obvious signs of recovery for this market.
So how do you see the impact of UAW strike and Tesla's aggressive price reductions?
Like, according to the U.S. customers, the strike has indeed had some impact on their on the order deployments of our products. But as I said, I think most automakers have now reached preliminary agreements with the union. And we also hope the strike will come to an end soon. I think the wage increases, I mean, on the U.S. automakers, I mean, it's going to put some cost pressure on their cost structure. Because they are not going to, they are not, I mean, just competing against each other and Tesla, but also the Japanese automakers and the Chinese automakers as well. So the cost structures of the Chinese players and the Japanese players can be very different from the European or the American automakers.
And then speaking of the Tesla, I mean, the Tesla's aggressive price reduction. I think Tesla's price reductions have put significant pressure on all automakers. So it will definitely affect the procurement prices for other OEMs as well. So as a supplier, we must continue to reduce our production costs to support our customers in lowering their relatively high prices of their vehicles.
So how do you see the impact of automakers adopting Tesla's charging standard? Has shipment based on the NACS resumed?
I think the case for the NACS charging standards is just a little bit similar to the charging standard, I mean, the charging speed of Type-C. So eventually, in order to make the charging environment more friendly, so the charging speeds are going to be universal. So I think that is good for the overall market to further increase the penetration rates of the electric vehicles with the availability of better or more friendly charging ecosystem.
And then what are the expansion plans for the new year? And what is the current capacity in Thailand and India?
We are following the previously mentioned plan for expansion. Currently, we have new factories under construction in Taiwan, Thailand, and the United States, Hungary, India, and Chongqing, China. I think they will contribute to capacity gradually. Thailand and India currently account for about 30% of the company's capacity, and there is still some room for further expansion. Just in the stage of expanding our footprints in many markets. So we're not only building up our capacity in Thailand, but also we are building up an R&D center in Thailand. And then we also have another R&D center in Bangalore in India. And then also in Europe, we also have a plan to build up a new R&D center there.
So it's not only about the capacity expansion, but also about the expansion of our global operations and footprints. I think if you look at the labor market in Taiwan, I mean, considering the shrinking population, it's actually getting more and more challenging to recruit enough engineers and good workers, good employees. So that's the reason why we will need to recruit the engineers, I mean, from other regions, not just limited to the engineers, I mean, in Taiwan.
Why was there an improvement in gross margin and operating margin in Q3 compared to Q2?
The improvement of Q3 in Q3 compared to Q2 was primarily due to an increase in the economics of scale and some minor product portfolio improvements. And then the next question is, how should we assume the future R&D expense ratio? So I think the guideline is it's highly likely that we will continue to invest at least 8% of our annual revenues into R&D every year. So after all, our commitment to R&D will remain, I mean, unwavering, and it is actually the source of our competitive competitiveness.
So what are the CapEx plans for next year and for this year?
So our CapEx for the first three quarters has already reached around TWD 21.2 billion. It is estimated the total CapEx for this year will increase from last year. So in the coming years, due to the constructions of several new plants and R&D centers, the need for machinery and equipment for expansion, the demand for automation, I mean, our CapEx is expected to remain above the TWD 20 billion mark.
So has the U.S. chip control on the Chinese market impacted your business?
I think, such conflicts do not, sorry, do have some indirect impact on our business, but not direct impact. So that is the reason why, I mean, there are so many companies, they are actually also building up their capacity overseas, not limited to China, because of the increasing geopolitical tensions.
So what about the building automation business? When do you expect to see improvement?
Well, building automation had good, I mean, actually had pretty decent growth in the first half of this year. Mainly because many projects with longer delivery times that were secured before the economic slowdown were still being shipped. However, we have seen, I mean, a clear slowdown in the market in the second half. So but, I mean, looking ahead, I think there are multiple drivers driving the secular growth of this market and this business. So, for example, the increasing requirements for energy efficiency and a healthier environment, and then also, the requirements for a smarter and smarter building automation management. So in long run, we are still quite optimistic about the growth, I mean, for this business. But just in the near term, there might be some headwinds for the business to cap the growth in the near term.
So can you comment on the profitability for the charging station business?
There are actually a wide range of charging products. So for example, for the entry level charging product, which is the AC chargers, the prices, I mean, of these AC chargers are significantly lower than the DC chargers, and also the market is more competitive. And what we are seeing in the market currently is, if you look at the China auto market, there are something called switching batteries vehicles. So in order to support this, I mean, the system of switching batteries, that you've got to have the fast chargers to charge the batteries fast enough.
So we do think, that, I mean, the trend of, I mean, more and more increasing numbers of switching battery cars, is going to be one of the tailwind for the penetration of the EV chargers in the market. So I think there are actually, I mean, there is actually plenty room for the further development and further upgrade of the charging of the EV charger products in the market. And then also there is also plenty room for further price reduction for the products as well.
So how has the performance of data center solutions been this year?
I think like, just like, many other businesses, this market has slowed down, I mean, this year, for the conventional or general servers. But we should still achieve a modest positive growth for this business.
So have you seen a restocking in the PC market? Does AI PCs, like AI servers, have higher power demands and that can drive the value of power supplies?
So to be honest, I don't really get the idea of AI PCs, what they are really for. Okay, I will have our CIO, [Lambert], to maybe to give us share his insights.
I think there are actually many emerging terms in the market, but in my perspective, I think the AI PCs are just like the extension of gaming PCs.
So, for the new segment, called mobility, will you start to disclose your financial numbers with these four new segments next quarter, from next quarter? And in... I mean, whether or not the infrastructure, I mean, the charging infrastructure business is part of, I mean, this mobility segment.
For a second question, the answer is no, because the EV charger is actually part of the infrastructure, so we keep it, I mean, we keep it within the infrastructure segment. So we only separate the onboard EV components from power electronics, and then we make it a new segment called mobility. I think this, I mean, new segment, we're going to take effect from the beginning of next year. So for Q4 revenues, we will continue to use the existing, I mean, the current business category to disclose our numbers. So we will start to disclose the numbers, I mean, with the new business segments from Q1 of next year.
So and then I have a question regarding the telecom power business. When do you expect to see a recovery for the telecom power business?
I think that last year is going to be better than this year.
So what is the sales contribution of your EV-related business for this year and for next year?
This year is around 12%, I think it's probably around 12%, and next year is going to be higher. And the next question is regarding to the sales contribution of the AI servers or AI-related products. So I think we have, I mean, explained this before. There isn't really a very clear or specific definition about the AI servers.
So our definition is, okay, despite the fact that, I mean, there isn't a really concrete definition for the AI servers, but we define the servers with the GPU as AI servers. So I think, for the AI servers or for the GPU-based servers, we not only provide the AC adapter, AC power supplies, but also provide the DC to DC power supplies to the AI servers. But for the sales contribution for next year, I think at this moment, because we are still in the process of finalizing the budget for next year. So what I can be sure is that next year is definitely to see, I mean, pretty significant growth. But, in terms of the sales contribution or as a percentage of our overall revenues, it's hard to predict. So are there any other questions?
So as you said, there was some impact, I mean, from the UAW strike. So as far as I know that, I mean, the EV business is turning to a profitable business this year. So, would there be any impact on the profitability because of this UAW event?
No, I think the profitability of the EV business is more related to the economics of scale. Despite the... I mean, I said, I mean, there was some impact on the business, I mean, by this, UAW strike, from this UAW strike. But because those, factories with the strikes are not really the factories making the electric vehicles, so the impact from the strikes is actually much, much less than the impact from the high interest rate environment.
So in the early days, the borrowing interest rates was around 1%-2%, but recently, it has already come to a level of 8%. So the high interest rate environment will still eventually, I mean, have some impact on the demand or and the macro, the macro demand. So I have a follow-up question on the EV business.
So, do you have any... Can you give us any guidance on the profitability or the margin profiles of the mobility segment?
I think the nature of the auto business, I mean, for everyone in this market, the margins that, I mean, are not going to be, they are not—I mean, this business is not going to be a high margin business anyway. So I think ideally, the margin, it should be somewhere around 15%.
But one—the reason why we separate this business and make it into a new segment, because we naturally want to make be more transparent, to, of course, our investors. I think I can provide you, I mean, give you an advice on the profitability forecast for this EV business. So I think that when you look at the margin profile for this mobility segment or this EV business, I think you can picture it as a contract manufacturing business. So if you look at the contract manufacturing guys, I mean, in Taiwan, even though their GP margins and OP margins are actually relatively low, but because they actually, in terms of the turnovers and overall scales of the businesses are, I mean, pretty large.
So still the returns and the profits on a dollar basis can still pretty be pretty decent for the investors. So I know there, I mean, there are many competitors in the market for the EV business, just like what we know, one of the Japanese companies, they actually have been always pretty ambitious about their goals.
So what do you see the competitive landscape in this market?
So I think that all, I mean, the only the big companies, I mean, they can really, they can really join the games, I mean, in this market. Because it is actually a one or zero market, which means that if you don't get the design win today, then you won't secure the orders and revenues in the next three years for a certain model. So, well, if you were the supplier for this EV market, in this EV market, I think that it not just requires pretty high capability and expertise in the product designs and manufacturing, but also requires pretty high level of your capital or financial preparation.
For example, once you get the design wins or the project wins from the customers, that before you are going to the stage of mass production, that you, before that, you need to, to get a qualification for your production line, for one year. And so the cost is already there before you can make the revenues and make money from the business. So this business nature and the business model is actually pretty challenging, for the companies. So, only the large companies can have this capability to handle this kind of, I mean, can handle this.
So can you please provide more colors on your thermal management products or solutions for AI servers, for, or for other ICT applications?
I think that we not only provide cooling products for the ICT space, but also we provide cooling solutions for the auto space as well. So we actually also provide the cooling products for the batteries on the vehicles. I think we actually are seeing an increasing sales contribution from the auto application for, I mean, within our cooling fan business.
So my question is, is the qualification process for the GPU-based server is longer than the conventional servers?
No, I think they are similar to other servers, because we have been always, I mean, producing the high powers. Yeah, because the basic technology for the higher power output products is actually similar. So I don't think there is any significant difference in terms of the qualification process.
Okay, so can you give us any updates with your lawsuit, I mean, with, against Vicor?
So because it's, I mean, still in the legal proceeding, so, so we are not allowed to, I mean, share any, comments or updates on this.
Do you have any target for your ideal debt ratio? Considering the current higher interest rate environment.
So yeah, I think the high interest rate indeed has, I mean, impact in many aspects. So that is also one of the reasons why we just sold some of our shares of Delta Thailand recently. Because I mean, selling the shares, I mean, of Delta Thailand, that we were able to pay down some of the debt. So we will continue to, I mean, invest in China. So, we actually are still...
I mean, despite many new overseas capacity, but we are still, I mean, in the process of, I mean, expanding our investments and capacity in China, in many different provinces. So we are not really moving away from China, but just overseas capacity outgrows, I mean, in terms of the pace of expansion, outgrows the capacity in China. So I think our goal, I mean, for the capacity, I mean, the China capacity as a percentage of overall capacity is probably near or around 50% eventually. And then, as I said, we are we are not really moving away from China.
Because if you look at the manufacturing environment in China, or the whole supply chain, is highly, highly efficient, and it is way more efficient than any other markets, even compared to Thailand. So that's the reason why we will continue to, I mean, invest in China. So we are just in the process of, I mean, diversify our global footprints. But it doesn't mean that we are, I mean, lowering our investments in China.
So I can be sure that the capacity, of overall capacity, I mean China capacity, is definitely going to be above 50% mark. So can you share the outlook for the China market?
So, well, we don't really have many exposure, direct, direct exposure to the China market. So I think in terms of the sales, China, I mean, sales to the China market is only below 15%.
So, thank you for coming today. Thank you. So, yeah.